Power distribution network cluster division method and system considering power coordination and source-load-storage matching

Abstract

The invention discloses a power distribution network cluster division method considering power coordination and source-load-storage matching, and relates to the field of optimized operation and management of a power system power distribution network, and the method comprises the following steps: introducing a source-load-storage matching degree, an electrical distance-based modularity, a cluster active balance degree, a cluster reactive balance degree, and a cluster energy storage supply and demand coordination index; and a comprehensive comprehensive performance optimization index system is constructed. According to the method, balance autonomy in the clusters is realized, and power interaction between the clusters is remarkably reduced. By means of the particle swarm optimization, the complex power distribution network cluster division problem can be efficiently solved, and the optimal cluster division scheme is obtained through optimization. According to the method, the operation efficiency and reliability of the power distribution network can be effectively improved, and powerful support is provided for intelligent and efficient management of the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of optimal operation and management of distribution networks in power systems. Specifically, it relates to a method and system for partitioning distribution network clusters considering power coordination and source-load-storage matching. Background Art

[0002] With the continuous increase in the penetration rate of renewable energy in power systems, how to effectively manage and dispatch these distributed energy sources has become an important research topic. Traditional power system management methods are mainly based on static network structures and determined load demands, and it is difficult to cope with the volatility and uncertainty of renewable energy generation. For example, wind and solar power generation are affected by weather and seasons, and their output power is not only unpredictable but also changes frequently. Traditional scheduling methods often lead to waste of resources and unstable operation of the system.

[0003] Distribution network cluster partitioning is an efficient method. By dividing the network into multiple autonomous sub-networks (clusters), better source-load-storage matching and voltage regulation can be achieved within each cluster. However, existing cluster partitioning methods often only consider structural indicators such as modularity, while ignoring functional indicators such as source-load matching degree, active power balance degree, reactive power balance degree, and energy storage supply-demand coordination indicators. This results in a low matching degree between renewable energy and loads within the cluster, and frequent power interaction with other clusters is required, increasing the complexity and operating cost of the system. Therefore, it is necessary to deeply study the method for partitioning distribution network clusters. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method for partitioning distribution network clusters considering power coordination and source-load-storage matching, including the following steps:

[0005] Step 1: Obtain the specific network structure of the distribution network, as well as the output powers, loads, and sizes of energy storage devices connected to each node.

[0006] Step 2: Define the modularity index, source-load-storage matching index, active power balance index, reactive power balance index, and energy storage supply-demand coordination index of the node system as the comprehensive performance optimization indexes for the cluster partitioning algorithm.

[0007] Step 3: Determine the constraint conditions involved in the distribution network cluster partitioning problem and perform modeling.

[0008] Step 4: Initialize the population and parameters of the particle swarm optimization algorithm, and determine the population size, dimension, number of iterations, individual learning factor, global learning factor, and inertia weight.

[0009] Step 5: Generate a number of particles by using the method of initializing the population with chaotic mapping. Each particle represents a clustering division method, and calculate the fitness of each particle by using the comprehensive performance index proposed in Step 2 as the fitness function;

[0010] Step 6: Sort the particle swarm according to the fitness obtained in Step 5, determine the particle with the best fitness as the global optimal particle. At the same time, each particle retains its individual optimal position;

[0011] Step 7: Update the velocity and position of the particles according to the update formulas of the velocity and position of the particle swarm, and calculate the fitness value for the new positions;

[0012] Step 8: Repeat Step 7 until the maximum number of iterations is reached or the fitness value no longer changes, and output the optimal distribution network clustering division scheme and the fitness value at this time.

[0013] Preferably, in the process of obtaining the comprehensive performance optimization index of the clustering division algorithm, the coordination degree between the net load in the cluster and the renewable energy output is evaluated through the source-load-storage matching degree, which is used to show the matching degree between the renewable energy output power of the cluster as a whole and the optimized energy consumption demand.

[0014] Preferably, in the process of obtaining the comprehensive performance optimization index of the clustering division algorithm, the structural strength of the clustering division is evaluated through the modularity index.

[0015] Preferably, in the process of obtaining the comprehensive performance optimization index of the clustering division algorithm, the active power balance index is expressed as:

[0016]

[0017] where, Ω x represents the x-th cluster in the cluster set {Ω1, Ω2, …, Ωx, …}, P sup，Ωx is the active power that can be provided by the distributed photovoltaic in the cluster Ω x , P need，Ωx is the active power demand in the cluster Ω x , and is the active power balance of the entire distribution network.

[0018] Preferably, in the process of obtaining the comprehensive performance optimization index of the clustering division algorithm, the reactive power balance index is expressed as:

[0019]

[0020] where, is the reactive power balance index of the cluster Ωx; is the reactive power that can be provided by the cluster Ωx, is the reactive power demand, is the reactive power balance degree of the entire distribution network.

[0021] Preferably, in the process of obtaining the comprehensive performance optimization index of the cluster division algorithm, the coordination index of energy storage supply and demand is used to reflect the supply and demand coordination between distributed energy storage and net load in the cluster division, and the coordination degree between the adjustable resources of the source, load and energy storage in each cluster and the system operation requirements.

[0022] Preferably, in the process of obtaining the comprehensive performance optimization index of the cluster division algorithm, based on five indexes, the multi-objective problem is converted into a single-objective problem by linear weighting, and the fitness deviation weight method is used to obtain the size of each weight.

[0023] Preferably, in the process of obtaining the constraint conditions, the distribution network power flow constraint, node voltage constraint, power balance constraint and distributed energy storage constraint are used as the constraint conditions.

[0024] Preferably, in the process of generating a number of particles by using the method of initializing the population with chaotic mapping, first, two chaotic variables m k,j and n k,j are generated; then, the generated chaotic variables are used to update the particle positions; finally, the fitness of x(D) and p(D) is compared and the larger particle is selected, where the sub-dimension is D, x max,j and x min,j are the maximum and minimum values of the particle search positions in the dimension respectively, and m 0 and n 0 are two arbitrary initialized chaotic variables.

[0025] The present invention discloses a distribution network cluster division system for realizing the distribution network cluster division method considering power coordination and source-load-energy storage matching, including:

[0026] A data acquisition module for obtaining the specific network structure of the distribution network, the output powers, loads and energy storage sizes of different resources connected to each node;

[0027] An index definition module for defining the modularity index, source-load-energy storage matching index, active power balance degree index, reactive power balance degree index and energy storage supply and demand coordination index of the node system as the comprehensive performance optimization index of the cluster division algorithm;

[0028] A model construction module for determining the constraint conditions involved in the distribution network cluster division problem and modeling;

[0029] An initialization module for initializing the particle swarm algorithm population and parameters, and determining the population size, dimension, iteration number, individual learning factor, global learning factor and inertia weight;

[0030] A particle generation module, which is used to generate a number of particles by using the method of initializing the population with chaotic mapping. Each particle represents a cluster partitioning method, and the fitness of each particle is calculated with the comprehensive performance index as the fitness function;

[0031] An optimization module, which sorts the particle swarm according to the calculated fitness, determines the particle with the best fitness as the global optimal particle. At the same time, each particle retains its individual optimal position;

[0032] An update module, which is used to update the velocity and position of the particles according to the update formulas of the velocity and position of the particle swarm, and calculate the fitness value of the new position until the maximum number of iterations is met or the fitness value no longer changes, and output the optimal distribution network cluster partitioning scheme and the fitness value at this time.

[0033] The present invention discloses the following technical effects:

[0034] The present invention constructs a comprehensive comprehensive performance optimization index system by introducing the source-load-storage matching degree, modularity based on electrical distance, cluster active power balance degree, cluster reactive power balance degree and cluster energy storage supply-demand coordination index. This method not only realizes the balance and autonomy within the cluster, but also significantly reduces the power interaction between clusters. Through the particle swarm algorithm, the present invention can efficiently solve the complex distribution network cluster partitioning problem and optimize to obtain the best cluster partitioning scheme. This method can effectively improve the operation efficiency and reliability of the distribution network and provide strong support for the intelligent and efficient management of the power grid. Description of the Drawings

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 It is a diagram of the improved IEEE33-node system described in the present invention;

[0037] Figure 2 It is a curve graph of the predicted output of wind power and photovoltaic power and the load described in the present invention;

[0038] Figure 3 It is a schematic diagram of the method flow described in the present invention. Detailed Embodiments

[0039] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Usually, the components of the embodiments of this application described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but only represents the selected embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of this application.

[0040] As Figures 1-3 shown, the present invention provides a method for partitioning a distribution network cluster that comprehensively considers structure and function, including the following steps:

[0041] Step 1: First, obtain the specific network structure of the distribution network, as well as the different resource outputs, loads, and the sizes of energy storage connected to each node.

[0042] Step 2: Define the calculation methods for the modularity index, source-load-storage matching index, active power balance index, reactive power balance index, and energy storage supply-demand coordination index of the node system. Comprehensively consider the above five indicators as the comprehensive performance optimization indicators for the cluster partitioning algorithm, and assign corresponding weights to the five indicators to transform the multi-objective problem into a single-objective problem.

[0043] Step 3: Determine the constraint conditions involved in the distribution network cluster partitioning problem and perform modeling.

[0044] Step 4: Initialize the population and parameters of the particle swarm algorithm, and determine the population size, dimension, number of iterations, individual learning factor, global learning factor, and inertia weight.

[0045] Step 5: Use the method of initializing the population with chaotic mapping to generate N particles, that is, N distribution network cluster partitioning schemes. Each particle represents a cluster partitioning method, and calculate the fitness of each particle using the comprehensive performance index proposed in Step 2 as the fitness function.

[0046] Step 6: Sort the particle swarm according to the fitness obtained in Step 5, and determine the particle with the best fitness as the global optimal particle P gbest , and at the same time, each particle retains its individual optimal position P pbest .

[0047] Step 7: Update the velocity and position of the particles according to the update formulas for the velocity and position of the particle swarm, and calculate the fitness value for the new position.

[0048] Step 8: Repeat Step 7 until the maximum number of iterations is reached or the fitness value no longer changes, and output the optimal distribution network clustering scheme and the fitness value at this time.

[0049] Specifically, in Step 1, the improved IEEE 33-node is adopted for analysis in the present invention, and the specific network structure is as shown in the appendix. Figure 1 Among them, photovoltaic power generation units are connected to nodes 13, 16, 20, 24, and 31, with each installed capacity of 500 kW; wind power generation units are connected to nodes 6, 26, and 29, with each installed capacity of 800 kW; energy storage devices with a rated capacity of 800 kW·h are connected to nodes 6, 24, and 31, with a maximum charge and discharge power of 150 kW, and the upper and lower limits of the state of charge are 0.9 and 0.1. Each node includes a fixed load, a shiftable load, and a curtailable load. The shiftable load accounts for 5% of the load, and the curtailable load accounts for 10% of the load. The total predicted output of wind power and photovoltaic power in the system and the load curve are as shown in the appendix. Figure 2 as shown.

[0050] In Step 2, the specific definitions and calculation methods of the five indicators and the comprehensive performance indicator are as follows:

[0051] Source-load-storage matching degree indicator:

[0052] Considering the uncertainty of wind power and photovoltaic power, the cluster must use flexible loads and energy storage devices to regulate power, so as to construct an optimization model of the net load. The optimization scope of the model covers all nodes in the cluster, and its core goal is to make the optimized load curve as close as possible to the power generation curve of renewable energy in the time dimension, that is, to minimize the net load within the cluster.

[0053]

[0054] In the formula: is the net load value after the optimized scheduling of the source-load-storage of the kth cluster at time t; is the load value after the optimized scheduling of cluster k at time t; is the total power value of renewable energy within cluster k at time t; is the original load power value of the ith node within cluster k at time t; is the regulation power value of all adjustable devices within cluster k at time t; is the load amount participated in the scheduling by the flexible load in node i; is the real-time power regulation ability of the energy storage device installed in node i; is the optimization goal of the net load optimization model.

[0055] To improve the energy utilization rate during the optimal operation stage of the distribution network and reduce cross-regional power interaction, the net load within the cluster after optimal scheduling is used as the basis for evaluating the matching degree of source, load, and storage. The matching degree of source, load, and storage is a quantitative indicator used to evaluate the coordination degree between the net load within the cluster and the output of renewable energy. This indicator shows the matching degree between the output power of renewable energy and the optimized energy demand of the cluster as a whole, and is defined as follows:

[0056]

[0057]

[0058] In the formula: is the interaction matching degree of source, load, and storage resources within the kth cluster at time t, and is the matching degree index of source, load, and storage for the entire distribution network; N Ω is the number of clusters; T is the scheduling period.

[0059] Modularity index:

[0060] Modularity is a parameter that measures the strength of the network structure and is related to the network structure and edge weights. The higher the modularity, the more similar the functions and properties between nodes in the same region, and the stronger the structure of the network. In the distribution network

[0061] the edge weights of the modularity network are represented by the electrical distance. Therefore, the modularity index based on the electrical distance can be used to evaluate the structural strength of cluster division.

[0062] In the distribution network, the electrical distance is obtained from the Jacobian matrix of the power flow calculation equation:

[0063]

[0064] In the formula, ΔP is the change in the active power injected into the node, and ΔQ is the change in the reactive power injected into the node; the Jacobian matrix composed of the block matrices J Pδ 、J PU 、J Qδ 、J QU represents the relationship between the change in active or reactive power and the voltage phase angle or amplitude; Δδ and ΔU are the changes in the node phase angle and amplitude, respectively. The above formula can be transformed into:

[0065]

[0066] In the formula, S δP and S δQ are the sensitivity matrices of the active and reactive power injected into the node and the voltage phase angle, and S UP and S UQ are the sensitivity matrices of the active and reactive power injected into the node and the voltage amplitude.

[0067] For a distribution network with N nodes, the voltage sensitivity S UQ,ij has an electrical distance of:

[0068]

[0069] In the formula, d ij is the electrical distance between node i and node j, S UQ,ij is the element in the i-th row and j-th column of the node reactive-power - voltage sensitivity matrix, and S UQ,jj is the element in the j-th row and j-th column of the reactive-power - voltage sensitivity matrix.

[0070] Based on the above calculations, the modularity index function based on electrical distance can be calculated as:

[0071]

[0072] In the formula, is the modularity; is the sum of all edge weights of the distribution network; A ij is the edge weight between two nodes i and j, and k i , k j respectively represent the sum of edge weights connected to nodes i and j. When nodes i and j are in the same cluster, δ(i,j) takes 1; when nodes i and j are not in the same cluster, δ(i,j) takes 0.

[0073] Active power balance index of the cluster:

[0074] To ensure the active power balance ability within the cluster, the calculation method of the active power balance degree of the cluster is defined as follows:

[0075]

[0076] In the formula, Ω x represents the x-th cluster in the cluster set {Ω1, Ω2, …, Ωx, …}, P sup，Ωx is the active power that distributed photovoltaics in cluster Ω x can provide, and P need，Ωx is the active power demand within cluster Ω x , is the active power balance degree of the entire distribution network.

[0077] Reactive power balance index of the cluster:

[0078] During the cluster division process, each cluster should also have a certain reactive power regulation ability. Therefore, the reactive power balance index is also defined:

[0079]

[0080] In the formula, is the reactive power balance index of cluster Ωx; is the reactive power that the cluster Ωx can provide, is the reactive power demand, is the reactive power balance of the entire distribution network.

[0081] Cluster energy storage supply-demand coordination index:

[0082] To reflect the supply-demand coordination between distributed energy storage and net load in cluster division, as well as the coordination degree between adjustable resources of source, load, and energy storage in each cluster and system operation requirements, the calculation method of the cluster energy storage supply-demand coordination index is defined as follows:

[0083]

[0084] In the formula, is the difference between the energy storage capacity and net power of cluster Ωx at time t.

[0085]

[0086]

[0087] In the formula, φ e is the energy storage supply-demand coordination index of cluster Ωx; is the overall energy storage supply-demand coordination index of the distribution network.

[0088] In the present invention, considering the above five indicators comprehensively, the problem of multi-objectives through linear weighting is transformed into a single-objective problem. The specific expression is as follows:

[0089]

[0090] In the formula, γ is the comprehensive performance index, and ω 1 , ω 2 , ω 3 , ω 4 , ω 5 are the weight values corresponding to the source-load-energy storage matching degree index, modularity index, cluster active power balance index, cluster reactive power balance index, and cluster energy storage supply-demand coordination index respectively. In the present invention, the fitness deviation weight method is used to obtain the size of each weight. The method for obtaining the size of the weight corresponding to each objective function is as follows:

[0091] Obtain the optimal solution x i ;

[0092] Substitute the optimal solutions of other objective functions except the i-th objective function into the i-th objective function, and calculate the fitness value f i ' of this objective function. The formula is as follows:

[0093]

[0094] Compare the optimal value and the fitness value of the \(i\)-th single-objective function, and calculate the corresponding standard deviation \(\Delta f\). i . The formula is as follows:

[0095]

[0096] The deviation is obtained by subtracting the fitness value from the maximum value of the objective function, and the result should always be greater than 0.

[0097] Calculate the weight coefficient \(\omega\) of the \(i\)-th objective function i , and ensure that the sum of all weight coefficients is 1. The formula is as follows:

[0098]

[0099] In step three, the main constraint conditions to be considered are as follows:

[0100] Distribution network power flow constraint:

[0101]

[0102] In the formula, \(P\) i,t , \(Q\) i,t are the active and reactive powers injected at node \(i\) at time \(t\); \(U\) i,t , \(U\) j,t are the voltage amplitudes of nodes \(i\) and \(j\) at time \(t\); \(G\) ij , \(B\) ij are the admittances between nodes \(i\) and \(j\), and \(\theta\) ij,t is the phase angle difference at time \(t\).

[0103] Node voltage constraint:

[0104]

[0105] In the formula, are the upper and lower limits of the voltage amplitude of node \(i\).

[0106] Power balance constraint:

[0107]

[0108] In the formula, \(N\) pcc , \(N\) Ω , \(N\) C are the number of branches connecting the bus and the cluster, the number of clusters, and the number of nodes in the cluster, respectively; are the active power of branch \(l\) connecting to the upper-level power grid and the network loss of line \(ij\) at time \(t\), respectively.

[0109] Distributed energy storage constraint:

[0110]

[0111] In the formula, P b,i,max is the maximum charge and discharge power of distributed energy storage at node i; is the state of charge of the energy storage at node i at time t; S b,i,max , S b,i,min are the upper and lower limits of the state of charge of the energy storage; η char , η disc are the charge and discharge powers of the energy storage at node i.

[0112] In step five, the specific process of the method for initializing the population using the chaotic mapping is as follows:

[0113] The basic idea of chaotic population initialization is to utilize the randomness and unpredictability in chaos theory to distribute the individuals of the initial population at multiple positions in the search space, so as to increase the search range of the solution space and thus have a greater probability of finding the global optimal solution. As the basic chaotic mapping, the Logistic mapping has good convergence performance and is easy to implement, and its mathematical expression is as follows:

[0114] x n+1 = μx n (1 - x n )(n = 0, 1, 2,...)

[0115] where μ is the control parameter. When the value of μ is taken as 4, the generated chaotic sequence is the most uniformly distributed. When x 0 ∈(0, 1), the chaotic variable x n ∈(0, 1).

[0116] Based on this, a population initialization method based on the Logistic chaotic mapping is designed to initialize the positions of the particles. Assume the particle dimension is D, x max,j and x min,j are respectively the maximum and minimum values of the particle in the D-dimensional search position, m 0 and n 0 are two arbitrary initialized chaotic variables. First, generate two chaotic variables m k,j and n k,j . Then, use the generated chaotic variables to update the particle positions. Finally, compare the fitness of x(D) and p(D) and select the particle with the larger value.

[0117] m k,j = μm k-1,j (1 - m k-1,j )

[0118] n k,j = μn k-1,j(1 - n k-1,j )

[0119] x i,j = x min,j + m k,j (x max,j - x min,j )

[0120] p i,j = p min,j + n k,j (p max,j - p min,j )

[0121] In step seven, the specific update formulas for the velocity and position of the mentioned particle swarm algorithm are as follows:

[0122] v id (t + 1)= ωv id (t)+ c 1 rand 1 (p id (t)- x id (t))+ c 2 rand 2 (g d (t)- x id (t))

[0123] x id (t + 1)= x id (t)+ v id (t + 1).

[0124] The present invention constructs a comprehensive comprehensive performance optimization index system by introducing source-load-storage matching degree, modularity based on electrical distance, cluster active power balance degree, cluster reactive power balance degree, and cluster energy storage supply-demand coordination index. This method not only realizes the balance and autonomy within the cluster, but also significantly reduces the power interaction between clusters. Through the particle swarm algorithm, the present invention can efficiently solve the complex distribution network cluster division problem and optimize the best cluster division scheme. This method can effectively improve the operation efficiency and reliability of the distribution network and provide strong support for the intelligent and efficient management of the power grid.

[0125] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate means for implementing the functions specified in one flow Figure 1 one flow or more flows and / or blocks Figure 1 or in a block or more blocks.

[0126] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0127] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A distribution network cluster division method considering power coordination and source-load-storage matching, characterized in that: The following steps are involved: Step 1: Obtain the specific network structure of the distribution network and the different resource outputs and loads connected to each node as well as the size of energy storage; Step 2: Define the modularity index, source-load-storage matching index, active power balance index, reactive power balance index, and energy storage supply-demand coordination index of the node system as comprehensive performance optimization indicators of the cluster partitioning algorithm; Step 3: Determine the constraints involved in the distribution network cluster partitioning problem and model it; Step 4: Initialize the particle swarm algorithm population and parameters, determine the population size, dimension, number of iterations, individual learning factor, global learning factor, and inertia weight; Step 5: Use the method of initializing the population using chaotic mapping to generate a number of particles, each particle represents a cluster division method, and use the comprehensive performance index proposed in step 2 as the fitness function to calculate the fitness of each particle; Step 6: Sort the particle swarm according to the fitness obtained in step 5, determine the particle with the best fitness as the global optimal particle, and at the same time, each particle retains its individual optimal position; Step 7: Update the speed and position of the particle according to the update formula of the speed and position of the particle group, and calculate the fitness value of the new position; Step 8: Repeat step 7 until the maximum number of iterations is met or the fitness value does not change, and output the optimal distribution network cluster division scheme and fitness value at this time.

2. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 1 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, the coordination degree between the net load and the renewable energy output in the cluster is evaluated through the source-load-storage matching degree, which is used to show the matching degree between the renewable energy output power of the cluster as a whole and the optimized energy demand.

3. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 2 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, the structural strength of the cluster partitioning is evaluated by the modularity index.

4. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 3 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, the active power balance index is expressed as: In the formula, Ω x represents the xth cluster in the set of clusters {Ω1, Ω2, …, Ωx, …}, P sup，Ωx For cluster Ω x The active power that distributed photovoltaics can provide, P need，Ωx For cluster Ω x The active power demand within It is the active power balance of the entire distribution network.

5. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 4 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, the reactive power balance index is expressed as: In the formula, For cluster Ω x Reactive power balance index; For cluster Ω x The reactive power available, is the reactive power demand, It is the reactive balance degree of the whole distribution network.

6. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 5 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, the energy storage supply and demand coordination index is used to reflect the supply and demand coordination of distributed energy storage and net load in cluster partitioning, as well as the coordination degree between the source, load and storage adjustable resources of each cluster and the system operation requirements.

7. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 6 is characterized in that: In the process of obtaining the comprehensive performance optimization index of the cluster partitioning algorithm, based on five indicators, the multi-objective linear weighted problem is converted into a single-objective problem, in which the fitness deviation weight method is used to obtain the size of each weight.

8. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 7 is characterized in that: In the process of obtaining the constraint conditions, the distribution network flow constraint, the node voltage constraint, the power balance constraint and the distributed energy storage constraint are used as the constraint conditions.

9. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to claim 8 is characterized in that: In the process of generating several particles by using the method of initializing the population with chaotic mapping, firstly, two chaotic variables m are generated. k,j and n k,j ; Then, the generated chaotic variables are used to update the particle positions; Finally, the fitness of x(D) and p(D) is compared and the larger particle is selected, where the sub-dimension is D, x max,j and x min,j are the maximum and minimum values ​​of the particle in the dimension search position, respectively, m0 and n0 are two arbitrary initialized chaotic variables.

10. The method for dividing distribution network clusters considering power coordination and source-load-storage matching according to any one of claims 1 to 9, characterized in that: A distribution network cluster division system considering power coordination and source-load-storage matching for realizing a distribution network cluster division method, comprising: The data acquisition module is used to obtain the specific network structure of the distribution network and the output and load of different resources connected to each node, as well as the size of energy storage; The indicator definition module is used to define the modularity indicator, source-load-storage matching indicator, active power balance indicator, reactive power balance indicator, and energy storage supply and demand coordination indicator of the node system as comprehensive performance optimization indicators of the cluster partitioning algorithm; A model building module is used to determine the constraints involved in the distribution network cluster partitioning problem and model it; Initialization module, used to initialize the particle swarm algorithm population and parameters, determine the population size, dimension, number of iterations, individual learning factor, global learning factor, and inertia weight; The particle generation module is used to generate a number of particles by using the method of initializing the population with chaotic mapping. Each particle represents a cluster division method, and the fitness of each particle is calculated using the comprehensive performance index as the fitness function; The optimization module sorts the particle swarm according to the calculated fitness, and determines the particle with the best fitness as the global optimal particle. At the same time, each particle retains its individual optimal position. The update module is used to update the speed and position of the particles according to the update formula of the speed and position of the particle group, and calculate the fitness value for the new position until the maximum number of iterations is met or the fitness value no longer changes, and output the optimal distribution network cluster division plan and fitness value at this time.

Images